Learn R Programming

DESP (version 0.2-2)

Estimation of Diagonal Elements of Sparse Precision-Matrices

Description

Several estimators of the diagonal elements of a sparse precision (inverse covariance) matrix from a sample of Gaussian vectors for a given matrix of estimated marginal regression coefficients. Moreover, a robust estimator of the precision matrix is proposed. To install package 'gurobi', instructions at and .

Copy Link

Version

Install

install.packages('DESP')

Monthly Downloads

45

Version

0.2-2

License

GPL-3

Maintainer

Samuel Balmand

Last Published

February 2nd, 2017

Functions in DESP (0.2-2)

DESP_MST

Estimation of DESP using minimum spanning trees
DESP_RML

Estimation of DESP by relaxed likelihood maximization
DESP_AD

Estimation of DESP by average absolute deviation around the mean
DESP_OLS_B

Estimation of B by ordinary least squares
DESP_PEN_grad

Steepest descent algorithm for penalized maximum likelihood estimation
DESP_PML

Estimation of DESP by penalized likelihood minimization
DESP_SPT_MaxDegreeRoot

SPT computation choosing root a priori as the node of maximal degree
DESP_SimpleOut

Detection of simple outliers
DESP_MST_MaxDegreeRoot

MST computation
DESP_RV

Estimation of DESP by residual variance
DESP_SPT_MaxWeight

Maximum weighted tree among all shortest path trees computation
DESP_SPT_MaxDegreeRoot2

SPT - of maximum height equal to 1 - computation choosing root a priori as the node of maximal degree
DESP_Weighted_Graph

Graph representation from the matrix B
DESP_SRL_B

Estimation of the coefficient matrix
desp

Robust estimation of a sparse precision matrix
scsSOCP

solve a second-order cone program using SCS
DESP_SPT

Estimation of DESP using shortest path trees
DESP_SqPartCorr

Squared partial correlations computation
desp.cv

Selection of the tuning parameters of desp by v-fold cross-validation
DESP-internal

Internal DESP Functions
sqR_Lasso

computation of beta that minimize |Y-X*beta|_2 + lambda |beta|_1 (square-root Lasso)